Paul Scott
Build a dissertation search strategy by turning your research question into concepts, identifying suitable databases and combining relevant terms. Then record exactly what you searched, where you searched and what you found.
The aim is a focused, repeatable search strategy that supports your argument without swallowing your writing time. Start with clear boundaries before collecting more papers.
Key Takeaways
- Let your research question determine your concepts, databases and inclusion criteria.
- Combine synonyms with OR, connect concepts with AND, and adapt the syntax for each database.
- Save your search strings and results counts, follow citations, and schedule an update before your final draft.
Define Your Dissertation Search Strategy’s Scope
Set Boundaries Before Collecting Papers
Before starting a literature search, write a short scope statement covering your population, setting, central concepts and relevant period. Also record exclusions, such as school pupils when your question concerns university students.
Begin with recent review articles to learn the vocabulary and debates. However, an empty search doesn’t prove that nobody has studied your subject. Different terminology, populations or measures can hide relevant evidence.
Your search should support the comparisons developed in your dissertation literature review structure, rather than produce a disconnected reading list.
Match the Search to the Review Type
A traditional dissertation literature review needs a transparent, justified literature search strategy. A formal systematic review requires more extensive planning and reproducibility.
UNC’s systematic review search guide reports an average of 44 hours for a systematic review team to conduct literature searches. That figure isn’t a timetable for every dissertation.
Its eight tasks are defining concepts, choosing databases, listing terms, adding controlled vocabulary and selecting field tags. The remaining tasks are combining terms with AND and OR, translating searches between databases and saving the strategy.
Choose Databases That Fit Your Discipline
Consult your university librarian before starting the main search. Bring your question and initial terms so they can recommend relevant databases and help prevent aimless searching.
These are useful starting points, rather than compulsory database combinations.
| Discipline | Subject Databases | Useful Additional Coverage |
|---|---|---|
| Education | ERIC, ProQuest Education Database | PsycINFO for psychological aspects |
| Health | PubMed, Embase, CINAHL | PsycINFO for mental health |
| Business | ABI/INFORM, Business Source Complete | Scopus or Web of Science |
| Humanities | MLA International Bibliography, Historical Abstracts | JSTOR and library catalogues |
The University of Queensland’s database selection guide also identifies ABI/INFORM, Business Source Complete and Web of Science for business-related reviews.
Record the platform too. Accessing PsycINFO through EBSCOhost or Ovid gives you the same database, but the platforms use different search commands.
Google Scholar helps locate citations and accessible versions. However, its changing results and limited search controls make it a supplement to a documented database search.
Turn Your Question Into Search Concepts
Take this education question: “How do university students experience belonging during online learning?”
It contains three searchable concepts: online learning, university students and belonging. You don’t need to search every word in the question.
For health questions, PICO (population, intervention, comparison and outcome) can help identify concepts, but its elements won’t fit every dissertation topic.
Build a vocabulary table before opening advanced search.
| Concept | Initial Search Terms |
|---|---|
| Online learning | “online learning”, “distance learning”, elearning |
| University students | “university students”, undergraduates, “higher education” |
| Belonging | belonging, connectedness, “social integration” |
These terms are starting points, not exact equivalents. For example, connectedness may describe something broader than belonging, so assess relevance during screening.
An empty search alone doesn’t establish a research gap. First check for alternative terminology, populations or measures.
Borrow additional language from useful abstracts and subject headings. Include alternative spellings, abbreviations and older terminology where these matter.
Keep inclusion criteria separate from search terms. You may need to assess study design or participant age during screening because abstracts don’t always report them consistently.
Combine Terms With Boolean Operators
Build a Complete Search String
OR broadens a search by joining synonyms. AND narrows it by requiring both concepts. NOT excludes records containing a term, although it can remove useful evidence.
For the education question, a starting string is:
("online learning" OR "distance learning" OR elearning) AND ("university students" OR undergraduates OR "higher education") AND (belonging OR connectedness OR "social integration")
Use parentheses to keep each synonym group together, and match every opening bracket with a closing bracket.
Test the Logic Before Adding Limits
UNC illustrates nesting with (influenza OR flu) AND vaccine. Without parentheses, influenza OR flu AND vaccine doesn’t reliably express the intended grouping.
A search such as "online learning" NOT "high school" excludes records mentioning high school. However, it could also remove a useful comparison between school and university education.
Inspect an initial batch of results. If relevant studies are missing, revisit terminology or remove an overly restrictive concept. If results are irrelevant, try more precise terms or fields.
Quotation marks can search an exact phrase, while a truncation symbol can retrieve word variants. For example, educat* may find education, educator and educational when the database supports that syntax. Check each database’s rules before translating these features, as quotation marks, truncation and proximity commands vary between platforms.
Add Controlled Vocabulary and Field Tags
Combine Keywords With Subject Headings
Keywords search authors’ wording. Controlled vocabulary uses standardised subject headings assigned to records, which can bring together papers using different terminology.
PubMed uses MeSH; Embase uses EMTREE; CINAHL uses CINAHL Headings; and PsycINFO uses the APA Thesaurus. The University of Toronto’s subject-heading guide explains these database differences.
On EBSCOhost, CINAHL uses MH for subject headings and MM for major headings, while the latter uses DE for descriptors.
Combine relevant headings with keyword alternatives using OR. Keywords remain important because newer records may lack completed subject indexing.
Restrict Fields Without Losing Relevant Studies
Field tags specify where a database should look. According to PubMed’s search documentation, [ti] searches titles, [tiab] searches titles and abstracts, and [au] searches authors. Its [tw] field covers text-word fields including titles, abstracts, keywords and MeSH terms.
For example, "distance education"[tiab] limits that phrase to titles and abstracts.
Title-only searches can miss relevant papers whose titles use broader language. Start with suitable title-and-abstract fields, then refine. Translate field tags, headings and proximity operators separately for every database and platform.
Keep a Search Log You Can Report
Create a Word or Excel table before running your main searches. Record each search version separately so you can distinguish exploratory searches from the final strategy.
Use these fields for each entry.
| Search Log Field | What to Record |
|---|---|
| Database and platform | For example, ERIC through EBSCOhost |
| Date searched | The actual search date |
| Exact search string | Full terms, brackets and field tags |
| Limits | Publication dates, languages and document types |
| Results | Number retrieved before deduplication |
| Notes | Changes made, reasons and export filename |
Save exports and the original strategies alongside the log. Zotero can organise references and identify duplicates, but imported bibliographic records still need correction.
In your methodology chapter, explain database choices, search dates, concepts, limits and selection criteria. Put lengthy database-specific strings in an appendix where appropriate. For a systematic review, PRISMA-S provides search-reporting guidance.
Describe what you actually did, including departures from your plan. A saved strategy and completed log are clearer milestones than “work on dissertation research”.
Follow Citations and Keep the Search Current
Trace References and Include Theses
Citation searching, also called cited reference searching, follows a source’s connections. Backward searching examines the bibliography of an article, book or chapter. Forward searching finds later works that cite it.
Use Google Scholar, Scopus or Web of Science for forward searches. Record the starting source and date.
Also consider grey literature, including dissertations, theses and technical reports. PsycINFO indexes dissertations and technical reports, whilst university repositories and ProQuest Dissertations & Theses offer further routes. Judge each source by its methods and relevance, not its publication type.
Schedule Updates and Set a Stopping Point
Set specific database or journal alerts, Google Scholar alerts, Google Alerts or RSS feeds. Narrow terms keep the incoming reading manageable.
Before finalising your draft, rerun the original strategies and remove duplicates. In PubMed, an entry-date filter using [EDAT] can retrieve records added since the previous search. Entry date differs from publication date.
For a traditional review, stop expanding once you’ve completed the planned coverage and citation checks, addressed important disagreements and scheduled the final update. Record remaining limitations. For a systematic review, follow the protocol’s coverage and update requirements.
Frequently Asked Questions About Dissertation Search Strategies
How Many Databases Should I Search?
There isn’t a universal number. Choose databases that cover your discipline and the different aspects of your question. Explain their contribution rather than adding databases to increase the count.
A librarian can help identify overlapping coverage and missing subject areas. Your department’s requirements and review type also shape the selection.
How Do I Decide Which Results Deserve Close Reading?
Screen titles and abstracts against your inclusion criteria, then read the full text of potentially relevant studies. Keep reasons for excluding sources at the full-text stage when your review method requires them.
For central papers, examine participants, measures, methods and limitations. Our guide to reading academic papers effectively helps turn that reading into useful comparative notes.
Build a Search You Can Explain
A strong dissertation search strategy connects your question to relevant evidence through clear, recorded decisions. Database choice, terminology and search limits should all have a reason.
Start with your scope statement, then use the search strategy to test a concept-based search and save the exact version you use. A manageable, documented search gives you a firmer basis for selecting evidence and writing your review.